Take home message

• Ley pastures can rebuild soil organic matter and have long-lasting benefits for subsequent soil function, nutrient supply and crop productivity
• Despite lower input costs during their growing period, pastures generally produced lower income and gross margins than the cropping system over the same period
• Crop sequences following the pasture phases regularly had reduced input costs (fertiliser or pesticides) and either similar or higher income
• Over the full 8-year period, the ley pasture systems had $0-340/ha/yr lower gross margin returns compared to continuous cropping, but input costs were reduced by 35-45%
• For each year of pasture grown, the GM deficit was about $30-60 per ha, but this was offset by the accumulation of soil organic matter (up to 500 kg C and 60-80 kg N per ha per yr)
• Pastures needed additional N inputs from fertiliser or accompanying legumes to increase biomass inputs, soil carbon and nutrients in the soil.

Why consider regenerative practices?

Northern farming systems have long been supported by highly fertile soils that have supplied the majority of nutrients to crops. However, as our cropping soils become ‘older’ we are starting to see the signs of declining soil function and fertility, putting increased pressure on costly external fertiliser inputs which are required to be well-matched to crop demand to optimise cost-effectiveness and efficiency and avoid environmental losses. It is well understood that soils with higher soil organic matter are more able to provide nutrients to crops when they are needed, providing valuable mineralised nutrients when the crop demand is highest. Soils with declining organic matter also face the risks of deteriorating soil structure reducing the efficient capture and availability of rainfall and reduced crop establishment. The challenge for our farming systems is how we might rectify, regenerate or maintain the productive capacity of our soils and thereby support the ongoing productivity and profitability of our cropping systems.

A range of so called ‘regenerative’ practices are known to have beneficial effects on building soil health and function (e.g. ley pastures, green manures, cover crops, organic amendments), but often the long-term benefits of these are hard to quantify or occur irregularly. Questions remain about the productivity and economic benefits or costs of implementing these systems compared to conventional cropping systems. In our northern farming systems projects we have implemented some systems that would be considered ‘regenerative’ and have collected information on their relative productivity, impacts on soil nutrient balances and organic matter, and potential for reducing external inputs and improving overall soil health.

Regenerative practices employed in farming systems experiments

Over the past 10 years, in addition to treatments involving altering grain crop choices, crop intensity or the approach to nutrient inputs, several experimental sites in the farming systems project have implemented phases of ley pastures in rotation with grain crops. Until now we have reported little about the impacts of these interventions on profitability, input use efficiencies, nutrient balances and cycling, soil pathogens, as well as the actual outcome of rebuilding the soils organic matter and health.

Ley pasture systems implemented

Three locations spanning quite different production environments have implemented ley pasture phases of varying length and type (Trangie, Billa Billa and Pampas). At the Trangie site in central west NSW, a lucerne pasture phase of 3 years was sown on two soil types (Red Kandosol and Grey Vertosol) in 2016 (Table 1). The lucerne stand on the grey soil was severely depleted by waterlogging in 2018 and terminated after 2 years. Both sites were then returned to a cropping phase for 4 crops before being resown to lucerne in 2024.

Table 1. Crop history of the conventional continuous cropping compared with a lucerne ley pasture system at Trangie sites between 2016 and 2024. X – indicates a fallow year with no winter crop sown.

At the two southern Queensland sites, Billa Billa and Pampas, a subtropical grass-based pasture was grown for two different longevities, a short phase of 3.5 years or a long phase of >8 years (Table 2). At Billa Billa, Bambatsi grass was established in summer 2016 and was maintained for either 3.5 years or 8 years before being returned to a cropping phase. At Pampas, a Bambatsi and Rhodes grass pasture was established in autumn 2015 and has either remained for the entirety of the following 9 years or was removed after 3.5 years and returned to cropping for 4 crops before being reestablished in spring 2023. At both sites these grass-based pastures were managed with either no additional fertiliser inputs or with additional inputs of urea each year (either 50 or 100 kg N/ha depending on the season). The Pampas site also included a grass-legume mixed pasture over the same periods, which included a summer- (burgundy bean) and winter-active legume (snail medic).

At all sites, pastures were sown at commercially recommended rates with starter fertiliser at establishment. Selective herbicides were used to control weeds prior to and during the establishment period, and then occasionally in subsequent years if there was an outbreak of a particular weed in those systems. Pastures were terminated using non-selective herbicides and no tillage, which often required 2-3 repeated applications. Plots then remained fallow until soil water conditions were satisfied to sow grain crops back into these systems.

The pastures were not grazed directly with livestock, but biomass was cut and removed up to 3 times each year. Cuts were made at mid-flowering in lucerne to a height of 5 cm and the tropical grass pastures were cut to approximately 15-20 cm height when active growing plants had reached maturity (i.e. seed was present) with the aim of removing approximately 40% of the available biomass. This management may not have maximised pasture productivity, but is likely to represent a reasonable approximation of what might be achieved under typical grazing management.

To replicate the recycling of nutrients of livestock grazing rather than the total removal that would occur from our cutting regime, at the start of each summer growing season we replaced 75% of the macro nutrients (N, P, K and S) that were removed in the previous year. This was done with a balanced mix of muriate of potash, DAP, sulphate of ammonia and urea; these replacement fertiliser was not included in our economic calculations.

Table 2. Crop history of the conventional continuous cropping system compared with the ley-pasture systems at long-term farming systems experiments at Billa Billa and Pampas between 2015 and 2024. X indicates a 6-month fallow period; thick lines indicate the period used for productivity calculations below.

Crop and ley pasture production and economic returns

Trangie sites

At both Trangie sites the lucerne pasture phase was estimated to have produced about 40-50% of the income of the cropping system over the same period (Table 3). Input costs of both fertilisers and pesticides were also substantially lower but this reduction was not sufficient to offset the lower returns. Ultimately the return per ha over this period (2016 to 2018) was lower from the ley pasture than the cropping system – this difference was much greater on the red soil ($850/ha) despite the higher lucerne productivity (Table 3).

During the subsequent cropping phase following the removal of the lucerne pastures, crop production continued with lower costs compared to the continuous cropping. Nutrient inputs were reduced on the Grey-soil site, and pesticide costs reduced on the Red-soil site. On the Red-soil site following crops also generated slightly higher income which contributed to $1000/ha greater income from this period compared to the continuous cropping. This effectively balanced the income deficit during the pasture phase so that both systems have been estimated to achieve very similar total GM over the whole 8 years. On the Grey soil site there was less crop income generated due to a difference in the crop sequence being applied over the following period – ultimately the GM on this soil was $100/ha/yr lower in the ley pasture system than the continuous cropping system.

Table 3. Production and gross margin (GM) comparison of a lucerne ley pasture system compared to a continuous cropping system at the Trangie farming systems experiments between 2016 and 2023. Crop income was based on long-term grain prices (Bell et al., 2023); pasture income was calculated using $0.137 per kg of removed biomass, based on an assumption of $50/DSE/yr (i.e. per 365 kg DM consumed) as reported for beef grazing systems.

Billa Billa site

The high initial N status at this site meant that there was limited response to added N in the pasture productivity until after year 3 in the pasture phase (only 1.3 t DM/ha more harvested) (Table 4). Over the same period, little nitrogen fertiliser was applied in the cropping system, but nonetheless the cropping system still had substantially higher input costs over the first 4 years. Over the first 4-year period the cropping system generated about $2000/ha higher gross margin than the grass-based pastures, due to the substantially higher income (Table 4).

Over the subsequent 4 years, once soil N reserves had been depleted, there was a significant response to the additions of fertiliser N in the pasture. Biomass removed was increased by 10 t DM/ha, which resulted in a net positive economic response to the additions of fertiliser in the pasture system of around $750/ha over the whole 8-year period (Table 4). Over this second 4-year period of pasture production, the fertilised pasture generated about $3000/ha in gross margin, which was slightly higher than the cropping system over this same period ($2740/ha); the unfertilised pasture was lower ($2060/ha).

Crops following the shorter-term pasture phase generated about $1200/ha higher gross margin than the continuous cropping system over the same period (Table 4). This was due to a small reduction in input costs, but mainly due to higher crop income from the crop sequence over that period. This recovered some of the lower returns obtained during the pasture phase but was not sufficient to make up the difference compared to the continuous cropping. At the end of 8 years, the system with the short pasture phase produced about $100/ha/yr less gross margin than the continuous cropping. The deficit of the long pasture phase was larger at about $240-340/ha/yr (Table 4).

Table 4. Production and gross margin (GM) comparison of a grass-based ley pasture system compared to a conventional cropping system at the Billa Billa farming systems experiment between 2015 and 2022. Crop income was based on long-term grain prices (Bell et al., 2023); pasture income was calculated using $0.137 per kg of removed biomass, based on an assumption of $50/DSE/yr (i.e. per 365 kg DM consumed) as reported for beef grazing systems.

Pampas site

As with Billa Billa, at Pampas the high background soil mineral N meant that additional N inputs or legumes resulted in little additional pasture production over the first 3 years (Table 5). In fact, the grass-legume pasture produced less biomass over this time. Due to high upfront fertiliser and seed costs of the pasture, and the limited fertiliser inputs required in the cropping system, input costs were similar to the cropping system over this period. The significantly lower income from the pasture over this period meant that the pastures generated >$2000/ha less gross margin than the continuous cropping over the same period (Table 5).

Again, similar to the Billa Billa site, once the background mineral N levels were depleted the advantage of the fertilised or legume-augmented pastures became evident. By the end of the 8 years of pasture there was a significant increase of 11t DM/ha of pasture harvested when additional N fertiliser was added; this generated a positive return on investment increasing total gross margin by over $1000/ha (Table 5). After 8 years, the grass-legume pasture produced about 4 t DM/ha and had an overall gross margin benefit of $400/ha compared to the unfertilised grass pasture, despite having a deficit of about $750/ha after the first 4 years.

The crops following the short pasture phase produced similar gross margins to the continuous cropping systems over the same period having lower crop inputs but not generating quite the same crop revenue (Table 5). Interestingly over this same period (from year 4 to year 8), the longer-term pastures generated similar or higher gross margins ($4650-5850/ha) to the systems under crop. However, over the whole 8-year period the ley pasture systems generated about $120-340/ha/yr lower gross margins than the continuous cropping systems.

Table 5. Production and gross margin (GM) comparison of a grass-based ley pasture system compared to a conventional cropping system at the Pampas farming systems experiment between 2015 and 2023. Crop income was based on long-term grain prices (Bell et al., 2023); pasture income was calculated using $0.137 per kg of removed biomass, based on an assumption of $50/DSE/yr (i.e. per 365 kg DM consumed) as reported for beef grazing systems.

Nutrient and carbon balance of ley pastures compared to continuous crop

Despite the direct cost of the ley pastures, one of the goals of this approach is to rebuild or maintain soil fertility and function compared to continuous cropping. At both Billa Billa and Pampas sites the ley pasture systems had a much-improved N balance compared to the continuous cropping system. The grass-only pastures with no additional fertiliser inputs still ran a negative N balance due to some removal of nutrients (as would happen in a grazed system), however, this was still substantially lower (about 30-50 kg N/ha/yr lower) than the cropping systems where nutrients were still being applied regularly. The pastures with N inputs from either legumes or fertilisers resulted in a net positive N balance, suggesting that the soil fertility and soil organic matter was increasing. In the case of the long-term fertilised pastures compared to the continuous cropping this difference was estimated to be over 500 kg N/ha at Pampas and over 900 kg N/ha at Billa Billa, over the 8 years. Once the short-term pastures were removed, the net positive N balance over the first 4 years was then depleted under the cropping phase as the release of mineral N offset fertiliser inputs required.

It is worth noting that our practice of replacing most of the nutrients removed from our cutting regime to simulate recycling from livestock was critical to maintain the nutrient balance of these systems. If the pastures had been managed for hay production, then the net removal in pasture biomass is similar to that from grain cropping and would similarly need to be replaced.

Table 6. Nitrogen balance under the grass-based pasture phases and subsequent crop sequences compared to the Baseline continuous cropping systems at Pampas and Billa Billa sites over 8 years.

^† - Nutrient replacement was calculated as 75% of removal to emulate nutrient cycling under grazing

Carbon inputs from growing plants are a key driver of changes in soil carbon. Table 7 below shows the estimated inputs of above- and below-ground plant residues from the continuous crop baseline and the ley pasture systems over the 8-year period of the experiments. This shows little advantage in terms of carbon inputs in the systems involving the shorter pasture phases. At Pampas and Trangie Grey soil sites, the continuous cropping systems were predicted to have more carbon inputs than those involving the 3-4 years of pasture followed by crops, while there was a small advantage of 4-7 t DM/ha at the Billa Billa and Trangie Red sites (Table 7). On the other hand, the longer-term grass-based systems involving a legume or fertiliser inputs were shown to have a significantly higher overall C input than the cropping systems. The grass-only pastures provided an additional 20-25 t DM/ha (8-10 t C/ha) inputs and the fertilised pastures an additional 40 t DM/ha (or 16 t C/ha) over the 8 years of pasture growth at the Billa Billa and Pampas sites.

Table 7. Crop residue inputs (t DM/ha) (above and below-ground) from continuous cropping and ley pasture systems at farming systems sites between 2015 and 2023. Below-ground biomass was estimated using root factors of 1.3 for annual crops and 2.0 for perennial pastures based on values quoted in the literature.

Soil carbon and function improvements under ley pastures

Corresponding to the higher estimated C inputs from the pastures, we have observed an increase in soil C and numerous other indicators of soil health over the experimental period (Table 8). For example, at Pampas the continuous cropping systems has been observed to maintain soil organic C levels, while under the pasture this has increased by 0.45%. Based on the soil bulk density at the site this is estimated to be a net increase of around 5 t soil C per hectare over this period. Soil total N has also increased correspondingly. Using the stoichiometric relationships for soil organic matter, these values also closely correspond with the N balance calculations shown in Table 6.

Notably the pasture ley has also enhanced soil microbial activity, and beneficial (free living) nematode numbers compared to the continuous cropping system. These are likely driven by the higher C inputs as a food source for microbes in the soil. However, similar changes and status was found for root lesion nematode (Pratylenchus thornei) and arbuscular mycorrhizal fungi (AMF) populations between the pastures and continuous cropping systems, probably owing to the crops grown in the baseline cropping system. In comparison to both these systems, an area of permanent fallow with limited or very little C inputs, shows how soil C and overall soil biology and health declines in the absence of plant residue inputs.

Table 8. Comparisons of some measures of soil organic matter and biology differences in the soil surface (0-10 cm) between the pasture ley, conventional cropping systems and the permanent fallow at Pampas. Letters denote significant differences between the systems for that attribute at that sampling time.

Conclusions

Ley pastures offer significant potential to increase soil carbon and improve overall soil health and function. As others have also shown, the higher carbon inputs and permanent cover provided by productive pastures can greatly enhance soil microbial activity and a range of beneficial microbiology. We have also demonstrated that ley pastures reduce the populations of damaging diseases and pathogens. Across all sites the ley pasture systems incurred significantly lower input costs both during the pasture phase as well as during the following cropping phase. However, these reductions in costs were rarely sufficient to offset the lower income that was estimated during the pasture phase. Using our data across sites and price assumptions here, we find that each year the pasture phase was maintained reduced the net gross margin return by about $30-60/ha/year.

The question is if the gains in soil fertility are sufficient over the long-term to offset this opportunity cost. Using our data here we estimate that the fertilised grass-based pastures were accumulating about 60-80 kg N/ha/year along with about 400-500 kg of soil C/ha/year. This accumulation of additional nitrogen ($80-100/ha/year) can have ongoing value for reducing N inputs and these are likely to last longer than the short periods measured here; meanwhile the continuous cropping system continued to run a negative nutrient balance which has not be accounted for in the gross margins here. Further benefits of soil C accumulation could have on soil structure, function or for greenhouse gas (GHG) mitigation credits could have further value to the farming system over the long-term.

Finally, it seems from our data that the fertiliser additions or legumes took more than 4 years to accrue their benefit to pasture productivity. Once soil mineral N reserves had been depleted and the pasture had tied-up any available mineral N, these additional inputs continued to maintain pasture productivity while the unfertilised pastures fell behind. Longer-term pastures also offered the greatest benefits for building soil health and mitigated the significant costs of establishment and termination that reduced the relative profitability of the shorter-term phases of pasture. While the gross margin returns of the pastures were lagging in the first 3-4 years compared to continuous cropping, at both sites the estimated returns from the established pastures was similar to the cropping systems over the subsequent years. These longer pasture phases are likely to be most beneficial to rebuilding soil and have a lower economic cost compared to shorter phases.

Acknowledgements

The research undertaken as part of this project is made possible by the significant contributions of growers through both trial cooperation and the support of the GRDC, DPIQ, NSW DPIRD and CSIRO, the authors would like to thank them for their continued support.

The Northern Farming Systems project has involved a large team of researchers and support staff, most of whom are not listed here – we thank them all for their input, hard work and dedication to executing these experiments over the past 9 years. We also acknowledge the various collaborators involved with collecting the experimental data, members of our reference groups for their input into the project, and farmer collaborators for hosting the farming systems experiments across the region.

References

Bell L, Whish, J and Horan H (2023) Short and long-term profitability of different farming systems - southern Qld, GRDC Grains Research Update paper, accessed January 2025.

Contact details

Lindsay Bell
CSIRO Toowoomba, QLD
Mb: 0409 881 988
Email: lindsay.bell@csiro.au

Date published
February 2025